A new study suggests mandated universal literacy screening programs aren’t translating into help for struggling young readers, including children with dyslexia.
Reading is one of the fundamentals of early education, but it’s also something millions of American kids struggle with. Forty percent of fourth graders have “below basic” reading skills—the worst figure in two decades, according to the National Assessment of Educational Progress.
To help get more kids on track for learning to read, most states now require universal literacy screening between kindergarten and third grade, aiming to spot problems early and direct resources to those needing extra help.
But a new nationwide survey of educators, led by a Boston University education researcher, suggests the mandates are failing, with struggling readers paying the price.
The study discovered a litany of barriers to successful screening implementation, from insufficient screener training to variation in scoring accuracy. Nearly half of the hundreds of educators surveyed across 39 states says their institutions had no systemic procedures in place for developing literacy programs.
According to the researchers, that means schools could be missing opportunities for early intervention, including with kids who have a learning disorder like dyslexia.
The study’s findings appear in the Annals of Dyslexia.
“In practice, it seems like the screening is a compliance exercise, rather than a tool to actually drive instructional modification on the ground,” says Ola Ozernov-Palchik, a BU Wheelock College of Education & Human Development research assistant professor of language and literacy education and expert on the cognitive neuroscience of language and literacy development. She helped lead the study, along with colleagues at MIT, Harvard University, and Florida State University.
“While educators overwhelmingly value screening, they reported limited training, inconsistent administration practices, and uncertainty about interpreting results.”
Here, Ozernov-Palchik explains what her team’s findings might mean for students, what needs to change, and how AI might help more kids learn to read:
What might the screening issues you uncovered mean for kids?
If you imagine a child is struggling with reading, before universal screening, that child would not be identified until third or fourth grade, when children in the United States are expected to know how to read and when the instruction in schools is transitioning from learning to read to reading to learn. We know that is way too late; at this point, many children may think they’re not smart, that teachers think that they’re just not trying hard enough. The gaps between them and their peers widen and become much more encompassing of other domains. They’re not motivated to go to school, and we know that closing those gaps becomes harder and harder with every year.
We know, based on our national literacy scores, we’re not doing a good job at teaching kids how to read—we have only 40% of children in our nation that read at a proficient level. Children who have dyslexia are less likely to graduate from high school and to go into higher education, are more likely to have a range of social-emotional difficulties, anxiety, depression.
But these screenings aren’t specifically just to identify kids who have dyslexia.
That’s correct. Universal screening is really to identify a risk, it’s an indicator that someone may not be developing as expected in the domain of literacy. It could be because of an indicator of dyslexia, an oral language issue, or another underlying reason. The idea is to flag that first risk in order to pay greater attention and do some follow-up, whether it’s additional instruction or assessments. If we’re not using the data to inform what we’re doing about the child, then there’s really no point. One of the educators says that they feel like screening is kind of this check-the-box exercise in the schools. There’s a mandate, we do this, but nothing changes on the ground.
Might kids from some backgrounds be affected more than others?
There were huge equity implications in our findings. For example, many educators responded that they did not know how to correctly score responses when the speakers were English language learners or children with a dialect. They were not trained on that. That’s a huge piece, because we know that these children are often underidentified in the context of screening and don’t receive the help they need. They have this double whammy—they’re learning the language and they’re not being identified if they have reading difficulties.
Is compliance a resource issue or is there some opposition to this screening that is affecting implementation?
Among the educators, there’s really broad support. Where it breaks down is implementation. Educators are saying, “We need more training.” For example, 75% of educators indicated they received fewer than three hours of training, 44% received less than one hour or no training at all. About 50% of the educators did not get a chance to practice before administering the screening to their first child. Only 40% of educators felt confident interpreting the screening data in order to drive their instructional decision or deciding what to do for their next step.
You recently spoke at the Massachusetts State House in support of a bill that would promote comprehensive literacy instruction in schools. What was the case you made for it?
We need an FDA-like process for education, just like if you go to the doctor, whatever you get from them is going to be evidence-based and rigorously validated. A similar approach needs to happen in education. Kids should not be spending hours and hours in school doing things that don’t have evidence for being effective. Our national trends in literacy really reflect that lack of evidence-based approaches in education. We have decades and decades of research on literacy; we know how to teach every single child how to read, so follow the evidence, require that schools use evidence-based curriculums.
Another piece that’s really important is data. What is happening with the data that’s being collected? It needs to be a process of accountability. This bill, for example, requires school districts to share their data with the Massachusetts Department of Elementary and Secondary Education as a lever of accountability, but also in order to realize where the needs are. If a particular school district is struggling, then that’s where you allocate the resources. There needs to be professional development, but that doesn’t end with a one-hour workshop; it needs to be continuous and systematic.
You’re studying how artificial intelligence (AI) could support children with language-based disabilities. What’s the potential there?
We’re launching a new initiative, Evidence-Based AI in Learning, as part of the BU AI and Education Initiative, and that’s a cross-disciplinary program between Wheelock and the Hariri Institute, with the idea of establishing a standard of evidence for educational technology that uses AI. I’m really passionate about how using AI can help us use data to transform education. AI is democratizing data analytics and data visualization across other domains, but it’s not really used in education as much. For example, teachers often feel unprepared to look at the graph of a child’s data and make instructional decisions based on that. Data visualization really focuses on how we can make this data story more apparent, more cohesive, and really drive decisions effectively and efficiently. We recently had a hackathon where we brought together educational technologists, AI people, researchers, educators to work around this problem.
Another domain is using AI to individualize instruction, so we’re developing a platform that uses a large language learning model to provide one-on-one tutoring for vocabulary in the context of books. Next, we’re going to start piloting the platform with kids and see if it drives their vocabulary scores.
What motivates you to do this work? Why do you care about literacy and literacy screening in particular?
I know we can teach kids how to read. Helping a child learn how to read is the most primary, most important function of our schools. If we can’t teach kids how to read, then what else? This is our number one job. We know how important literacy is for every single aspect of life: educational success is a huge predictor of health, of civic engagement. If we want to make our world a better one, we need to raise kids who know how to read, know how to comprehend, and can learn from reading and be critical consumers of information that they read.
This research was supported by the Chan Zuckerberg Initiative–funded Reach Every Reader project.
Source: Boston University
